A global survey of eating preferences, dietary habits and food beliefs

The development of eating preferences and behaviours is a complex multifactorial process (see Box 1). The food–medicine interface, especially in Asian medical systems such as Chinese medicine and Ayurveda, also influences food habits (World Health Organisation, 2003). As a consequence, many theories have been developed to explain eating preferences (Furst et al, 1996; Glanz et al, 1998; Levine et al, 2003).

Maslow’s Hierarchy of Needs theory depicts a pyramid with five levels of needs ranging from physiological need at the base of the pyramid progressing through safety needs, love/belonging needs and esteem needs to the need for self-actualisation at the highest level (Maslow, 1943). Lowenberg adapted Maslow’s theory to incorporate the different stages in the development of food habits (1970). The lowest level of the hierarchy applies to food eaten to survive, the next level to food eaten for social security, progressing to eating to belong, then eating according to one’s social status, and finally eating for self-actualisation. Lowenberg’s adaptation of Maslow’s theory might partly explain the differences in global eating preferences: people in countries with a limited food supply are likely to consume food to survive, while those in countries with an abundance of varied food are more likely to consume foods to satisfy higher personal needs.

In order to motivate people to make healthy food choices, it is important for health professionals to understand the factors that influence food preferences, habits, and beliefs. Many health professionals have difficulty motivating people from different cultural backgrounds to change their eating habits and sometimes regard culture as a barrier to effective care (Farooqi et al, 2000; Galasso et al, 2005).

The IDF DECS developed a multidisciplinary core curriculum and supporting teaching modules to educate health professionals to deliver diabetes education, including nutrition advice (IDF, 2006). In order to ensure the nutrition module was culturally relevant and appropriate, the authors undertook an exploratory study to determine global dietary habits on behalf of IDF DECS.

Aims of the study
The aims of the study were as follows.

1. To explore global eating trends and dietary habits. 
2. To identify similarities and differences among countries and regions.
3. To include the information in the IDF DECS nutrition module to suggest ways health professionals could advise people with diabetes about appropriate nutrition.

Materials and methods
A one-shot cross-sectional survey was used to collect the data using anonymous questionnaires in 2005. Questionnaires were emailed to a purposive sample of 55 health professionals (doctors, nurses and dietitians) providing diabetes care across the seven IDF regions. The sample comprised 35 individual health professionals and 20 health professionals working for various diabetes associations. There are some limitations to this method as purposive samples may not be representative of the general sampling population and so the range and source of the responses may not reflect those of all health professionals or people with diabetes in that area.

Sampling procedure
The IDF member associations were asked to forward the questionnaire to relevant health professionals in their respective regions. IDF DECS members were also asked to provide the names and contact details of health professionals in their regions who were likely to provide relevant information (n=55). One reminder email was sent to all potential participants three weeks after the questionnaire was first distributed to increase the response rate.

Questionnaire
The questionnaire consisted of 12 questions and asked about cultural influences on eating habits; staple foods; common carbohydrate, sugar, and fat foods; regional macronutrient recommendations; unhealthy eating habits; food belief; foods commonly used as medicine; and festivals and rituals (see Box 2).

The questionnaire was piloted by four health professionals in India and three IDF DECS members to ensure the content was relevant to the study aims, the language was appropriate and to determine whether the proposed distribution method was feasible before the questionnaire was distributed; however, it was not formally validated. Return of the questionnaire was taken as consent to participate and there was no direct working relationship between the authors and any of the participants who took part in the study.

Data analysis
The data were analysed using descriptive statistics and content analysis using the framework method. The framework method consists of a five-step process that involves a number of researchers becoming familiar with the data by reading it several times, establishing a thematic framework, indexing and charting emerging themes, and mapping and interpreting the findings (Ritchie and Spencer, 1994). Key words were identified in each question and the responses to each question were examined to identify the frequency with which the key word appeared. To understand the basis of any differences in food preferences and determine whether the preferences reflected food availability and economic factors, responses were divided into two groups defined according to the gross domestic product (GDP) at purchasing power parity (PPP) per capita. Developed countries were defined as having a GDP PPP per capita more than $20000 and developing countries as having a GDP PPP per capita less than $20000 (International Monetary Fund, 2006).

Results
Response rate
Forty-four health professionals from 36 countries responded giving a response rate of 81%. Some large countries with diverse ethnic backgrounds provided two or more responses, which were all included. Three respondents were excluded. One returned but did not complete the questionnaire, a second respondent from the South and Central America region (SACA) provided information for 16 SACA countries in one questionnaire, and the third provided information for all the Caribbean island countries in a single response. These responses were difficult to interpret and were excluded. Thus, 41 questionnaires from 34 countries were analysed.

Food preferences
The common carbohydrate, fat and sugar sources for developed and developing countries can be seen in Table 1.

Unhealthy foods
Unhealthy food was defined as being rich in fat, sugar or salt. The most common unhealthy foods are shown in Figures 1aand 1b. Overall, foods rich in fats, oils and sugar were the most commonly consumed unhealthy foods. Developed countries were more likely to consume snacks frequently, eat large portion sizes and high-calorie foods, whereas developing countries reported a low intake of fruits and vegetables and a high intake of starchy foods. Portion sizes and frequency were assessed by health professional’s judgement.

Common dietary beliefs
Thirty-nine health professionals provided information about common dietary beliefs of their patients. The responses were varied and contradictory suggesting food beliefs are a complex issue. The most commonly occurring belief was that fruit intake should be restricted. Other common beliefs were to restrict rice, starchy vegetables and carbohydrate-rich foods. It should be noted that the common food beliefs reported in this study are those health professionals encounter in their day-to-day contact with people with diabetes and may not represent the actual beliefs of people with diabetes in the area.

There was a general trend to recommend restricting sugar intake, especially in developed countries whereas health professionals in developing countries suggested that honey or molasses should be substituted for white sugar (see Figure 2). Likewise, opinions differed about artificial sweeteners. Some respondents suggested that ‘eating artificial sweeteners is fine,’ while others advised restricting them.

Food as medicine
Six respondents from developed countries and 26 from developing countries responded to the question about food eaten as medicine. Eighteen named bitter gourd (Momordica charantia) as a common food eaten as medicine in their area. Other foods eaten as medicine reported by developing countries were: Fenugreek (Trigonella foenum-graecum), Neem (Azadirachta indica), blackberry, traditional herbs and pumpkin.

Discussion
Some findings, such as health professionals’ perceptions about high intake of sugar and fat, were expected. Others, such as the low intake of fruits and vegetables in developing countries, were unexpected and could reflect the high cost of these foods in these countries, which supports the contention that food choices and eating preferences are partly influenced by the economic resources of individuals and countries. 

Although cultural factors play an important role in food choice, it appears that these factors are increasingly being overshadowed by globalisation, increasing urbanisation, agriculture and economics that improve food availability and influence choice – especially in developed countries (Shepherd, 1999; Drewnoski and Darmon, 2005; Drewnowski, 2007).

Soft drinks and beverages were major sources of sugar in both developed and developing countries. High sugar consumption has negative health effects such as increased prevalence of dental caries and high rates of peridontal disease in people with diabetes and hyperglycemia (Ismail et al, 1997; Mealey, 2006).

All respondents agreed that a high intake of sugar- and fat-rich foods was unhealthy. Unhealthy eating habits, especially high sugar and high fat diets, combined with increased periods of inactivity, are major contributors to the increasing prevalence of non-communicable diseases – especially obesity and diabetes (Popkin, 2006; Bray, 2004; National Institutes of Health, 2004).

Traditional medical systems such as Ayurveda and Chinese medicine recommend medicinal foods as a diabetes management strategy and are widely practised in developing countries. Other researchers have also reported Ayurvedic medicines being used to reduce blood glucose, including medicinal foods (Subbulakshmi and Naik, 2001).

The increased popularity of Chinese medicine and Ayurveda outside of Asia has triggered numerous studies to determine the hypoglycaemic effects of these medicines and foods on type 2 diabetes (Leatherdale et al, 1981; Ahmad et al, 1999; Krawinkel and Keding, 2006). Some foods have been shown to have hypoglycaemic properties, but more long-term studies are needed (Yeh et al, 2003; Kuriyan et al, 2007).

Cultural differences significantly affect eating habits among people from different regions in the study. Freimer et al (1983) suggested that, in order to improve concordance, it is important to consider food beliefs and practices before intervening: this suggestion has been incorporated into the IDF DECS curriculum (IDF, 2006).

Study limitations
Although responses were obtained from many parts of the world, the sample size was small and may not be representative or generalisable.  The questionnaire only had face and content validity and further validation is recommended before it is used in other studies.

Conclusion
Eating preferences differ between developed and developing countries. High intakes of fat, sugar, calorie-dense and processed foods, large portion sizes, and frequent snacking were common in developed countries. Low intake of fruits and vegetables and high intake of cereals and starch occurred in developing countries. Some foods were used for their medicinal properties, especially bitter gourd. This study reinforces the differences in eating preferences among people from different regions. In a multicultural society where people from different regions coexist, it is important that health professionals consider and understand traditional foods beliefs and practices and identify acceptable and accessible healthy food options before starting any form of intervention.

Acknowledgements
We wish to thank Marg McGill, Anne Belton, and Dr Sheridan Waldron from IDF DECS for their support in planning and conducting the survey. We wish to also thank all those who took part in the survey by distributing or responding to the questionnaires.